How independent these versions are. Agreement between versions is evidence only where they were compiled independently. Claude wrote first and drafted the brief. ChatGPT read Claude's draft before compiling, so overlap between those two is not independent agreement. Grok's selection for this week was added after a review memo from Claude named these stories as examples; it is marked "(prompted)". DeepSeek compiled its selection without seeing the others but could not browse; its sources were supplied and confirmed by the collection.
Stories across the versions:
| Story | Selected by |
|---|---|
| AI Diffusion Rule | Claude, ChatGPT, DeepSeek, Grok (prompted) |
| Executive order on AI infrastructure on federal land | Claude, ChatGPT |
| UK AI Opportunities Action Plan | Claude, ChatGPT |
| ChatGPT scheduled tasks | Claude, ChatGPT |
| MiniMax-01 | Claude, ChatGPT |
| Apple pauses AI news summaries | DeepSeek |
Claude (Claude Opus 5.5)
1. Biden Administration Issues the AI Diffusion Rule on Chip Exports
Date: January 13, 2025
Context: The Commerce Department released a framework for global exports of advanced AI chips. It set up three tiers:
- unrestricted access for 18 close allies;
- country caps on computing power for most of the world;
- effective bans for arms-embargoed countries such as China and Russia.
The rule also added controls on the weights of the most powerful closed models. It was published in the Federal Register on January 15 as an interim final rule, effective January 13, with most compliance requirements deferred to May 15.
Significance: It was the most ambitious attempt yet to govern where frontier compute could be built. It was issued a week before a change of administration.
Feedback: In TechTarget's January 13 report, Nvidia's Ned Finkle called the rule a "sweeping overreach." ITIF's Daniel Castro called it a further "misstep" that risked alienating partners. Some analysts quoted in the same report questioned its design.
Sources:
Hindsight: on May 13, 2025, the Commerce Department announced it would rescind the rule and told enforcement officials not to enforce it, so its May 15 compliance requirements would not be enforced. Source: BIS[link]
2. Executive Order Opens Federal Land to AI Data Centers
Date: January 14, 2025
Context: President Biden directed the Departments of Defense and Energy to lease federal sites for privately built AI data centers. Developers would have to bring their own clean-energy generation and pay for the facilities.
Significance: It addressed the main bottleneck of power and permitting, and treated frontier AI infrastructure as a matter of national security.
Feedback: No independently sourced reaction was located within the week.
Sources:
3. UK Launches the AI Opportunities Action Plan
Date: January 13, 2025
Context: Prime Minister Keir Starmer adopted all 50 recommendations of a plan led by entrepreneur Matt Clifford. They included:
- expanding the national AI Research Resource twentyfold by 2030;
- creating "AI Growth Zones" with faster planning and power access, starting at Culham in Oxfordshire.
Significance: Britain repositioned itself from AI safety convener to pro-growth adopter.
Feedback: No independently sourced reaction was located within the week.
Sources:
4. ChatGPT Gets "Tasks," Its First Scheduled Actions
Date: January 14, 2025
Context: OpenAI added a beta feature letting ChatGPT schedule reminders and recurring actions, such as a daily news briefing, for Plus, Team and Pro subscribers.
Significance: It was a small step, but the first time ChatGPT could act on its own schedule rather than only respond.
Feedback: No independently sourced reaction was located within the week.
Sources:
5. MiniMax Open-Sources MiniMax-01 with a 4-Million-Token Context
Date: January 15, 2025
Context: Shanghai-based MiniMax released the weights of MiniMax-01. It is a 456-billion-parameter mixture-of-experts model, with about 46 billion parameters active, built on "lightning attention," a linear-attention design. The company said it supports contexts of up to 4 million tokens.
Significance: Alongside DeepSeek-V3 from late December, it showed Chinese labs competing through architectural innovation and open weights.
Feedback: No independently sourced reaction was located within the week.
Sources:
ChatGPT (GPT-6)
1. The United States announces a broader AI export-control framework
Date: January 13, 2025 — effective date; January 15 — Federal Register publication.
Context: The Biden administration’s AI Diffusion Rule set out controls on advanced computing items and certain advanced AI model weights. Its effective date, publication date and compliance deadlines were different: most relevant compliance requirements were deferred until May 15. The framework sought to shape access to powerful AI infrastructure beyond existing restrictions focused on particular destinations.
Significance: The rule made international access to computing capacity a central policy issue for AI developers and their customers. Its details mattered as much as the headline announcement because implementation was phased.
Feedback: Nvidia’s Ned Finkle opposed the rule in a January 13 statement, arguing that it threatened innovation and economic growth. Nvidia was an affected supplier, not a disinterested observer.
Sources:
- Federal Register, January 15, primary rule and dates: federalregister.gov[link]
- Nvidia, January 13, primary industry response: blogs.nvidia.com[link]
2. Britain adopts the AI Opportunities Action Plan
Date: January 13, 2025 — publication and government response.
Context: The UK government endorsed the recommendations of Matt Clifford’s AI Opportunities Action Plan. Its response proposed expanded public computing capacity, AI Growth Zones and wider use of AI in public services. The commitments were a program of work, not evidence that the promised capacity or productivity gains had already materialized.
Significance: The plan treated adoption, infrastructure and public-sector capability as connected policy problems. It offered a national AI agenda that extended beyond developing a single leading model.
Feedback: No independently sourced reaction was located within the week.
Sources:
- UK government, January 13, primary action plan: gov.uk[link]
- UK government, January 13, official response: gov.uk[link]
3. ChatGPT adds scheduled tasks in beta
Date: January 14, 2025 — rollout announcement.
Context: OpenAI began rolling out scheduled tasks to ChatGPT Plus, Pro and Team users in an early beta. Users could request one-time reminders or recurring actions for a future time. The release introduced scheduling without establishing that ChatGPT could autonomously perform every kind of external task.
Significance: Scheduling changed the interaction from an immediate exchange into a continuing request. It was a concrete, bounded step toward assistants that remain useful after the user closes a conversation.
Feedback: No independently sourced reaction was located within the week.
Sources:
- OpenAI, primary release notes, January 14 entry: help.openai.com[link]
- MacRumors, January 14, contemporary coverage: macrumors.com[link]
4. MiniMax releases models built for unusually long contexts
Date: January 15, 2025 — model announcement; technical paper submitted January 14.
Context: MiniMax introduced MiniMax-Text-01 and MiniMax-VL-01, combining its Lightning Attention approach with a mixture-of-experts architecture. The team reported training with contexts up to one million tokens and extrapolating to four million during inference. These were the developers’ technical results, not independent proof of perfect recall across documents of that length.
Significance: The release put long-context efficiency alongside benchmark scores as a competitive dimension. Large document collections and extended interactions made the cost and reliability of retaining context practical concerns.
Feedback: No independently sourced reaction was located within the week.
Sources:
- MiniMax, January 15, primary release: minimax.io[link]
- MiniMax, technical paper submitted January 14: arxiv.org[link]
5. Biden orders federal support for AI infrastructure development
Date: January 14, 2025 — executive order signed; January 17 — publication.
Context: President Biden signed Executive Order 14141 on AI infrastructure. It directed federal agencies to support development of large AI data centers and associated energy infrastructure, including opportunities on federal land. The order combined expansion of computing capacity with conditions concerning security and clean energy.
Significance: The order treated energy and sites as constraints on AI development that federal policy could address. Model competition therefore had a land-use and power-supply dimension as well as a software dimension.
Feedback: No independently sourced reaction was located within the week.
Sources:
- Federal Register, January 17, primary executive order: federalregister.gov[link]
- American Presidency Project, January 14 document transcription: presidency.ucsb.edu[link]
DeepSeek V4.1-Flash-1017
1. Biden administration imposes sweeping new AI chip export restrictions
Date: Event date, January 13, 2025.
Context: The US Commerce Department announced an "AI diffusion rule" restricting exports of advanced AI chips and model weights to a wide range of countries, while exempting 18 allies. The rule took effect on January 13, with compliance due May 15.
Significance: The measure was the Biden administration's last major attempt to shape global AI compute flows before leaving office, and it immediately created uncertainty for chipmakers and cloud providers reliant on international sales.
Feedback: Nvidia publicly criticised the rule as "misguided" in a statement issued January 13, 2025.
Sources:
2. Apple temporarily disables AI news-summary feature after inaccuracies
Date: Report date, January 16–17, 2025.
Context: After Apple Intelligence's notification summaries repeatedly misrepresented news stories, including a false BBC News summary claiming that shooting suspect Luigi Mangione had shot himself, Apple disabled summaries for news and entertainment apps in the iOS 18.3 developer beta.
Significance: The retreat highlighted the reputational and editorial risks of deploying generative AI in news distribution, and gave news organisations a concrete example of AI-induced error at scale.
Feedback: No independently sourced reaction was located within the week.
Sources:
3–5. Slots left empty.
Grok 4
1. U.S. Commerce Department (BIS) publishes Framework for Artificial Intelligence Diffusion (AI Diffusion Rule) (prompted)
Date: January 13–15, 2025 (announcement January 13; Federal Register publication January 15)
Context: The Bureau of Industry and Security published an interim final rule establishing a global licensing framework for advanced computing integrated circuits and the model weights of the most advanced closed-weight AI models (those trained on 10^26 or more computational operations). It created a tiered system of destinations, new license exceptions (including Artificial Intelligence Authorization), updates to the Validated End-User program for data centers, and related controls intended to manage diffusion of advanced AI technology while protecting national security. Public comment period was opened.
Significance: It represented a major expansion of U.S. export controls specifically targeting both AI chips and frontier closed model weights on a worldwide basis.
Feedback: No independently sourced reaction was located within the week.
Sources: